INDEX

Explanations

reported speech or quotations within text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

speech attribution

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

parsedMessage

-0.67

تقاوى

-0.62

SourceChecksum

-0.60

mpto

-0.57

codable

-0.57

Vidite

-0.56

 surla

-0.54

webElementXpaths

-0.54

thâu

-0.53

zette

-0.53

POSITIVE LOGITS

 frecuentemente

0.36

 informaciones

0.35

 beroep

0.33

 rodillas

0.33

 aclarar

0.32

 sapore

0.32

ísimo

0.31

 preghiera

0.31

 beauté

0.31

 eski

0.31

Activations Density 0.019%

reported speech or quotations within text

speech attribution

No Comments

No Known Activations